DOI: http://dx.doi.org/10.26483/ijarcs.v9i1.5266 Volume 9, No. 1, January-February 2018 International Journal of Advanced Research in Computer Science REVIEW ARTICLE Available Online at www.ijarcs.info © 2015-19, IJARCS All Rights Reserved 285 ISSN No. 0976-5697 A STUDY OF MACHINE TRANSLATION APPROACHES FOR GUJARATI LANGUAGE Jatin C. Modh Assistant Professor, Narmada College of Computer Application, Bharuch, Gujarat, India. Dr. Jatinderkumar R. Saini Professor & I/C Director, Narmada College of Computer Application, Bharuch, Gujarat, India. Abstract: India is a multi-lingual country. At present, there are 22 official languages in India. Gujarat is a state located in the western region of India. The Gujarati language is spoken by nearly 60 million people worldwide, making it the 26th most-spoken native language in the world. In Machine Translation System (MTS), one natural language gets translated to another language using computational applications with minimal human effort or without a real-time human interface. Many attempts have been done in Machine Translation System for Indian languages. Unfortunately, we do not have an efficient Machine Translation System today. This paper gives a brief description of approaches of Machine Translation and the work done for the Gujarati language. Keywords: Machine Translation System (MTS); Computational Linguistics; English; Gujarati; Natural Language Processing I. INTRODUCTION Machine Translation [1] refers to the automated translation of text from one language to another language. Machine Translation System (MTS) is the application of Natural Language Processing (NLP) of Artificial Intelligence. The language of text entered as an input is known as the source language whereas the language of output text is known as the target language. Nowadays Machine Translation System is an emerging area of study for researchers in India. India is multilingual country. Indian government uses Hindi or English language as a communication medium whereas various states of India use their local language as a communication medium. There is a big demand for document conversion from one language to another language. The English language is widely used in all fields. So Machine Translation Systems are needed for translation of local language to English language or vice-a- versa. The Gujarat language is the official language of the state of Gujarat of India. Indian government publishes and issues official documents in English or Hindi or in both the languages. State government publishes official documents in their regional languages also. Gujarat Government uses the Gujarati language for official documents. In the Gujarat state, local newspapers, magazines and books are published in the local Gujarati language only. For the exchange of information among states, central government, industry, academia, good Machine Translation System (MTS) is required. Manual translation of documents is very time consuming and costly. This paper presents the approaches of Machine Translation and the work done for Machine Translation for Gujarati-English or English-Gujarati language pairs. II. OVERVIEW OF MACHINE TRANSLATION APPROACHES Researchers proposed many approaches for the Machine Translation. Overview of main approaches is presented here. There are two broad categories of Machine Translation Systems, namely Rule-Based and Empirical Based Machine Translation Systems. Hybrid Machine Translation system takes the benefits from both Rule-Based Machine Translation System and Empirical Based Machine Translation System. Rule-Based Machine Translation System is further classified into Direct, Transfer and Interlingua, while Empirical Based Translation System is classified into Statistical and Example- based machine translation system. Figure 1. Classification of Machine Translation System A. Rule-Based Machine Translation (RBMT) Rule-Based Machine Translation is a traditional method of Machine Translation and also known as Knowledge-Based Machine Translation [12]. RBMT uses grammar rules which Hybrid Machine Translation Statistical (Corpus) Based Example- Based Machine Translation Transfe r Interlingu a Direct Rule-Based Machine Translation Empirical Based Machine Translation